|Publication number||US7356232 B1|
|Application number||US 11/497,223|
|Publication date||Apr 8, 2008|
|Filing date||Aug 1, 2006|
|Priority date||Aug 1, 2006|
|Also published as||CN101173997A, CN101173997B, EP1884808A2, EP1884808A3, EP1884808B1, US20080069506|
|Publication number||11497223, 497223, US 7356232 B1, US 7356232B1, US-B1-7356232, US7356232 B1, US7356232B1|
|Inventors||David J DiGiovanni, Jayesh Jasapara, Andrew D. Yablon|
|Original Assignee||Furukawa Electric North America|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (4), Non-Patent Citations (1), Classifications (9), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This invention relates to optical fibers that are specially designed to carry high optical power.
At very high powers (approximately >1 MW) an optical beam increases the refractive index of the silica glass by an amount proportional to the local intensity. This can create a lens that focuses the beam down to a small spot (or a collection of small spots) causing optical damage. This process is termed self-focusing and has been extensively studied over the past several decades. (See R. W. Boyd, Nonlinear Optics, 2nd. Edition, Academic Press, Boston, 2003.) Self-focusing serves as an upper limit on the maximum power that can be guided in materials such as silica glass, and this limit is often termed the critical power for self focusing, Pcrit, which according to Fibich et al. (G. Fibich and A. L. Gaeta, “Critical power for self-focusing in bulk media and in hollow waveguides,” Optics Letters, vol. 25, pp. 335-337 (2000)) can be estimated as:
where λ is the wavelength, n0 is the material's original refractive index, and n2 is the nonlinear refractive index expressed in units of m2/W such that the total refractive index, n, is given by n=n0+n2l where l is the local optical intensity in W/m2. Since its refractive index, n0, is about 1.45 and its nonlinear refractive index, n2, is about 3×10−20 m2/W, at the commonly used laser wavelength of 1060 nm, Pcrit in silica glass is approximately equal to 3.8 MW. In other words, a bulk sample of silica glass will not be effective in guiding a 1060 nm optical beam at a power greater than about 3.8 MW because any light beam above this threshold power will rapidly focus to an infinitesimal spot size and damage the glass.
Equation (1) was thought to be valid for optical fibers as well as for homogeneous silica glass samples, thereby setting an upper limit on the peak power carrying capacity of any silica optical fiber. However, it turns out that certain specific optical fiber designs may suppress the inception of self-focusing and thereby permit such optical fibers to carry an optical signal with a power greater than Eq. (1) without self-focusing to the point of material damage.
Self-focusing is not the only limit to the power carried by optical fibers. Other important constraints include self-phase modulation, stimulated Brillouin scattering (SBS), stimulated Raman scattering (SRS), and dielectric optical breakdown. Unlike self-focusing, where the threshold is related to the total peak optical power, the threshold for these limitations as well as dielectric optical breakdown depends on the peak local intensity in the fiber. In general, avoidance of these limitations and dielectric optical breakdown is accomplished by increasing the effective modal area, Aeff, of the fiber. The effective area of an optical fiber mode is defined by Aeff=(∫|E|2dA)2/∫|E|4dA where E is the local electric field and the integrations are understood to be performed over the cross sectional area of the fiber. By increasing the effective modal area, the threshold power for intensity-dependent limitations can be elevated so that self-focusing becomes relatively more important. Calculations reveal that for near-infrared wavelengths and for reasonable Aeff (<3000 μm2), dielectric optical breakdown occurs at a lower power level than self-focusing when the pulse duration is about 1 ns or longer (including continuous wave signals). Since the present invention is designed to delay the onset of self-focusing, but is not designed to mitigate dielectric optical breakdown, it is expected to be most useful for fibers carrying high peak-power in pulse duration shorter than about 1 ns. The invention described here is intended for applications in which the peak optical power exceeds the conventional bulk-media self-focusing threshold. When the optical pulse length is relatively short, it is understood that the peak optical power will exceed the conventional bulk-media self-focusing threshold even though the average optical power integrated over a long time period may be well below the conventional bulk-media self-focusing threshold.
Distributing optical energy evenly across a large core optical fiber is one approach to maximizing the power capacity of the optical fiber without exceeding a threshold value of intensity-dependent limitations at any given point in the fiber. It is known that this can be achieved by engineering the refractive index profile of the optical fiber to produce a modefield with a flattened intensity profile as demonstrated in U.S. Patent Application Publication No. 2004/0247272. However it does not address the problem of self-focusing described above.
We have developed optical fiber designs that at least partly overcome the problems just described, mainly the problem of self-induced damage to optical fibers due to excessive self-focusing. The core refractive index of these fiber designs is grossly non-uniform in the center core of the optical fiber. In one embodiment, the optical fiber is designed with a deliberate and steep core trench. In addition, the nominal core region of these optical fibers has a very large area. The combination of these two properties restricts a large portion of the optical power envelope to a core ring, with reduced optical power inside the core ring. These designs substantially reduce self-focusing in the optical fiber. Photonic systems employing optical fibers having these modified core designs are expected to be especially effective for transmitting high peak optical power, e.g., greater than 1 MW, with short pulse duration.
However, as discussed above, even large core fibers like that represented by the profile of
Although the optical fiber profile of
Considering the problem from simple ray-optic theory, if a focus occurs then this means that the optical path length of the marginal ray approaching such a focus is the same as the optical path length of a ray on the optical axis. We make the rough approximation that the ray on the optical axis experiences a refractive index that is equal to n0+n2×I (where the nominal linear index of the glass is n0, and n2 is the non-linear component, i.e. the coefficient for the index delta created by on-axis high intensity I). It is also assumed that the marginal ray experiences a refractive index of only n0 (since the intensity is lower there). Since the optical paths of these two rays must be equal (optical path along optical axis=optical path for marginal ray), and using geometry, this equality becomes:
(n 0 +n 2 ×I)×L=n 0 ×L/cos(θ) (2)
where L is the physical distance along the optical axis to the focus and θ is the angle of the marginal ray relative to the optical axis.
Using algebra (and a simple approximation for cos(θ) the angle θ can be expressed as:
θ=(2×n 2 ×I/n 0)1/2 (3)
Assuming the original beam had transverse beam width 2×w0 (in this case 2×w0 is like the mode field diameter) then the marginal ray must travel a transverse distance of w0 to reach the optical axis. Accordingly, the approximate distance for self focusing, L, can be computed as:
L=w 0×(n 0/(2×n 2 ×I))1/2 (4)
All of these relations are approximate and are entirely based on a ray-optic picture of self-focusing that is valid well above the threshold for self-focusing Near the threshold for self-focusing, the convergence angle due to self-focusing (θ) will approximately balance the diverging angle due to diffraction of the beam (γ):
γ=λ/(n 0 ×w 0×π) (5)
where we have assumed that the transverse shape of the beam is approximately Gaussian. When the beam is approximately Gaussian in shape, the total power in the beam may be approximated as:
P=(π×I×w 0 2)/2 (6)
Equations (3), (5), and (6) may be algebraically combined to show that when the angle of diffraction is balanced by the converging angle due to self-focusing (γ=θ) then:
P=λ 2/(4n 2 πn 0) (7)
which is within a factor of 2 of the more exact solution presented in Equation (1). At higher powers the beam convergence due to self-focusing overpowers diffraction and the beam comes to a focus. At powers lower than Equation (7), diffraction overwhelms any convergence of the beam.
From the foregoing analysis it is concluded that the design goal for overcoming the self-focusing problem is to create a significant minimum in the optical intensity profile at the center of the optical fiber.
The optical fiber profile in the example shown in
where |E| is the electric field amplitude. Numerical simulations have shown that optical fibers satisfying these two conditions, i.e., a core with a large effective area, and a central region having a local minimum in the optical intensity, exhibit an elevated self-focusing threshold.
In some cases the presence of a minimum in the electric field can inhibit efficient optical coupling between the present invention and more conventional optical fibers exhibiting Gaussian-like mode field shapes. This can be overcome with the aid of mode field shape or size converting elements including bulk optical elements such as lenses or mirrors, as well as fiber based strategies such as those disclosed in US Provisional Patent Application Windeler and Yablon, Filed Dec. 16, 2005.
It should be understood that the minimum in the electric field described by equation (8) relates to the fundamental LP01 mode propagating in the core region of the fiber. However, it should also be understood that the invention is applicable to cases wherein the signal mode is LP02 or another mode.
As is evident from the above, the minimum in the electric field amplitude that is expressed by equation (8) can be obtained by producing a local minimum in the refractive index at the center of the core. In the usual case that minimum is less than the effective index of the fundamental mode. The width of the refractive index minimum also influences the depth of the minimum in the electric filed amplitude. The width of the refractive index trench will normally be larger than λ/2n, where λ is the vacuum wavelength and n is the refractive index.
The depressed region 22 in the profile of
Another approach to addressing the self-focusing problem is represented by the profile of
To demonstrate the stability of the fundamental propagating mode at high power in the optical fiber designs of the invention the variation of guided optical power of the fundamental LP01 mode with normalized modal effective index was calculated. The normalized modal effective index is given by neff-neff 0 where neff is the effective index of the mode when it is carrying high optical power (P˜Pcrit or P>Pcrit) and neff 0 is the effective index of the same mode at low powers (P<<Pcrit). The normalized modal effective index indicates the extent to which elevated optical power has perturbed the local refractive index. The effective index of a mode is defined by neff=βλ/(2π) where the phase of the optical mode accumulates at a rate β radians per unit distance.
It is evident from line 41 in
A concern for any numerical simulation of the fibers described here, when operating near or above Pcrit is the stability of the guided modes in the presence of small perturbations. If the guided modes are unstable, then any infinitesimal perturbation such as a small variation in refractive index profile may lead to catastrophic self-focusing and damage to the fiber. The stability of a fundamental LP01, is preserved for small perturbations if dP/dneff is greater than zero. See Kivshar et al., Optical Solutions: From Fibers to Photonic Crystals, Academic Press, New York, 2003. The optical energy carried by the fiber represented by curves 42 and 43 in
The electric field amplitudes can be computed for the various modes of the fiber using conventional numerical mode solving algorithms well known by those skilled in the art. The efficacy of a proposed fiber design can be evaluated by computing the desired signal mode at the operating wavelength based on the proposed index profile and the electric field amplitude (or optical intensity) can be checked to verify that a local minimum is present at the center of the fiber.
A simplified physical explanation for the efficacy of the present invention is as follows. In an axisymmetric waveguide such as a typical optical fiber, catastrophic self-focusing will normally occur when most of the optical signal focuses itself into the very center of the axis of symmetry (here, the very center of the fiber). Following the teachings described here, the refractive index at the center of the fiber is deliberately set low enough that the optical energy is inhibited from entering the center region of the fiber. Although some fraction of the optical energy will penetrate into the central depressed-index region of the fiber core (region 22 in
It is understood that as the optical power approaches and exceeds the bulk media critical power, Pcrit, the effective area Aeff of the signal mode will be reduced because of the intensity-induced perturbations to the index profile of the fiber. However, if the Aeff of the fiber at low power (P<<Pcrit) is high-enough, then this reduction in Aeff will not be sufficient to permit intensity dependent nonlinearities, such as dielectric optical breakdown, to impair the fiber.
While the high-power regime that is addressed here in terms of reduced self-focusing is mentioned earlier as above 1 MW, the power level where the effect begins to set in can be more precisely prescribed in terms of equation (7), which can be expressed in terms of an optical signal source wherein the optical signal has peak optical power greater than λ2/(4πn0n2) anywhere within the optical fiber, where λ is the vacuum wavelength, n0 is the unperturbed (linear) refractive index, and n2 is the nonlinear refractive index expressed in units of m2/W.
The terms up-doped and down-doped as used herein are terms well known to those skilled in the art. An up-doped glass or glass region is one that is doped to have a refractive index greater than that of pure silica. A down-doped glass or glass region is one that is doped to have a refractive index less than that of pure silica. Typically the host material is silica.
In concluding the detailed description, it should be noted that it will be obvious to those skilled in the art that many variations and modifications may be made to the preferred embodiment without substantial departure from the principles of the present invention. All such variations, modifications and equivalents are intended to be included herein as being within the scope of the present invention, as set forth in the claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5963700 *||Feb 25, 1998||Oct 5, 1999||Nippon Telegraph And Telephone Corporation||Optical fiber|
|US6018533 *||Sep 27, 1996||Jan 25, 2000||Ceramoptec Industries, Inc.||Optical fiber and integrated optic lasers with enhanced output power|
|US7130514 *||Jun 15, 2005||Oct 31, 2006||Corning Incorporated||High SBS threshold optical fiber|
|US20040247272||Sep 30, 2003||Dec 9, 2004||The Regents Of The University Of California||Flattened mode cylindrical and ribbon fibers and amplifiers|
|1||Hadley et al., "Bent-waveguide modeling of large-mode-area, double-clad fibers for high-power lasers" Paper No. 6102-63 Proceedings of the SPIE vol. 6102, Fiber Lasers III: Technology, Systems, and Applications, Published by SPIE (Bellingham, Washington) 2006.|
|Cooperative Classification||G02B6/4296, G02B6/02019, G02B6/02009, G02B6/032, G02B6/03611|
|European Classification||G02B6/02A2, G02B6/036H2|
|Aug 1, 2006||AS||Assignment|
Owner name: FURUKAWA ELECTRIC NORTH AMERICA, GEORGIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DIGIOVANNI, DAVID J.;JASPARA, JAYESH;YABLON, ANDREW D.;REEL/FRAME:018151/0629
Effective date: 20060801
|May 8, 2009||AS||Assignment|
Owner name: OFS FITEL, LLC., GEORGIA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:FURUKAWA ELECTRIC NORTH AMERICA, INC;REEL/FRAME:022668/0382
Effective date: 20081229
|May 19, 2009||AS||Assignment|
Owner name: FURUKAWA ELECTRIC CO., LTD., JAPAN
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:OFS FITEL, LLC;REEL/FRAME:022694/0729
Effective date: 20090331
|Sep 7, 2011||FPAY||Fee payment|
Year of fee payment: 4